Stated without spin — these are where a contributor's work most changes the trajectory. Seven invitations:

Cross-region atomic execution — no trustless solution yet; today, mitigation + economic underwriting (containment).
DA at GB/s with genuinely decentralized sampling — unproven at scale.
Prover supply and the energy of large-scale proving — a heavy, open cost.
Whether gravity-based placement converges to low cross-chain traffic — an unproven hypothesis.
The Asset Registry as a critical trusted component — it must be rock-solid.
Keeping the Global Root minimal so it never becomes a governance bottleneck.
Relayer/watcher economics — an incentive model that keeps liveness guaranteed.

Takeaway: XChain does not need to be perfect today; it needs to point the right way, stay honest about the path, and ship working pieces on a steady beat. These seven are open, and they are yours to take.